Probiotic effects are specific to the strain, not the genus. That one fact does more work than anything else in this field: a trial of Lacticaseibacillus rhamnosus GG tells you about that organism and nothing reliable about the other lactobacilli in the same capsule. The evidence is genuinely decent for a handful of narrow uses, mostly around antibiotic-associated diarrhea. It is thin to absent for routine use by healthy adults, and for “gut health” as a general proposition.
The version of this post that stood here claimed probiotic cultures produce a natural antibiotic effect, prevent infection by named pathogens, and manufacture vitamin B12 in the gut. None of those claims should have run, and they are corrected below.
Why the strain is the product
The working definition, from the FAO and WHO, is live microorganisms that, when administered in adequate amounts, confer a health benefit on the host. The operative words are “adequate amounts” and “a health benefit” — both of which have to be established for a specific organism, in a specific population, at a specific dose.
A full designation has three parts. Genus, species, strain: Lacticaseibacillus rhamnosus GG, or Saccharomyces boulardii CNCM I-745, or Bifidobacterium animalis subsp. lactis BB-12. A label that says only “Lactobacillus acidophilus” has told you roughly as much as a bottle labeled “dog.”
Naming got harder in 2020, when a taxonomic revision split the genus Lactobacillus into 25 genera. Familiar organisms were renamed: L. rhamnosus became Lacticaseibacillus rhamnosus, L. plantarum became Lactiplantibacillus plantarum. The strain codes did not change, which is why they are the part worth reading.
Where the evidence holds up
Antibiotic-associated diarrhea is the strongest case. A 2021 systematic review and meta-analysis in BMJ Open pooled 42 trials and 11,305 participants and reported a relative risk of 0.63 (95% CI 0.54 to 0.73) when probiotics were given alongside antibiotics, with moderate certainty by GRADE. The authors noted the effect was concentrated in populations at moderate-to-high baseline risk and may not carry over to lower-risk settings.
For Clostridioides difficile-associated diarrhea specifically, the Cochrane review updated in September 2025 found 31 trials and 8,672 participants, with incidence of 1.5 percent in probiotic groups against 4.0 percent in controls — a relative risk of 0.40 (95% CI 0.30 to 0.52), moderate-certainty evidence, number needed to treat 65. The reviewers were careful to add that the two largest studies in the analysis did not show a clear benefit.
On IBS the picture is mixed. Meta-analyses have found improvements in global symptoms, pain and bloating, but the trials are small and use dozens of different organisms, so the pooled estimate describes no product anyone can buy. The NCCIH summary records that the American College of Gastroenterology’s 2021 guideline recommended against probiotics for global IBS symptoms, citing a very low level of evidence. The 2020 American Gastroenterological Association guideline reached similar conclusions across the board, making conditional recommendations only for preventing necrotizing enterocolitis in preterm infants and C. difficile-associated diarrhea in patients on antibiotics, and none for Crohn’s disease, ulcerative colitis, pouchitis or IBS.
Where the evidence collapsed
Acute infectious diarrhea is the case worth studying, because the answer changed. Earlier editions of the Cochrane review indicated an effect. The 2020 update, covering 82 studies and 12,127 participants, reversed that. Restricting the analysis to trials at low risk of bias showed no difference between probiotic and control groups in the risk of diarrhea lasting 48 hours or longer, and the certainty of evidence on duration was rated very low. The reviewers attributed the earlier positive signal largely to publication bias: small positive studies were more likely to get published.
That is what a field correcting itself looks like, and a useful warning about every other probiotic claim resting on a pile of small trials.
The healthy-adult case is weaker still. A systematic review in Genome Medicine in 2016 examined seven randomized controlled trials in healthy adults and found no effect on fecal microbiota diversity, richness or evenness compared with placebo, concluding there was a lack of evidence for any impact on fecal microbiota composition in this group.
CFU counts, survivability, and what a label should say
Colony-forming units are the standard potency unit and are routinely misused as a proxy for quality. The NIH Office of Dietary Supplements states plainly that higher-CFU products are not necessarily more effective, and that typical products contain 1 to 10 billion CFU per dose. The dose that matters is the one used in the trial for that strain.
Survivability is the other half. Live organisms die in storage, faster when warm and moist, and gastric acid and bile kill more on the way through. That is why ISAPP, following the FAO/WHO guidelines, says a label should state the minimum viable count of each strain at the end of shelf life, not at manufacture — along with the full strain designation, a serving size that delivers the studied dose, and storage conditions.
| What a label says | What it actually tells you |
|---|---|
| “50 billion CFU” | Potency only. Not which organisms, not when measured, not whether that dose was studied |
| “Lactobacillus acidophilus” | A species. Different strains within it behave differently |
| “Lacticaseibacillus rhamnosus GG, 10 billion CFU at end of shelf life” | Enough to look up the trials and check the dose |
| “Proprietary probiotic blend” | Nothing verifiable |
Where all of this is weak
Almost every probiotic trial is small, and small trials in a commercially sponsored field are exactly the conditions in which publication bias thrives. The acute diarrhea reversal above is the demonstration, not the exception.
Pooling makes it worse rather than better. A meta-analysis combining twenty organisms at ten doses produces a number that applies to no actual product, and the guideline bodies keep saying so: the AGA noted its own recommendations lacked strain specificity because the underlying trials did not report it consistently.
Duration of effect is poorly studied too. Most supplemented strains are transient — they pass through, and colonization does not persist once dosing stops. Whether the measured effects outlast the capsule is largely unknown.
Safety, which is not a formality
For healthy adults, probiotics are generally well tolerated, gas and bloating being the common complaints. For seriously ill people the picture is different.
PROPATRIA, published in The Lancet in 2008, randomized 296 patients with predicted severe acute pancreatitis to a multispecies probiotic preparation or placebo. Infectious complications, the primary endpoint, did not differ. Mortality was 16 percent in the probiotic group against 6 percent in placebo, and eight patients in the probiotic arm developed fatal bowel ischemia against none in the placebo arm.
Fungemia is a documented risk with yeast probiotics. A study in the CDC’s Emerging Infectious Diseases in 2021 reviewed 46 cases of Saccharomyces fungemia across five Finnish university hospitals and found an odds ratio of 14 for probiotic use compared with controls. The authors concluded these products are not recommended for patients with indwelling catheters or who are immunocompromised or critically ill.
And in 2023 the FDA raised concerns about probiotic products used in hospitalized preterm infants after an infant death and more than two dozen adverse event reports, issuing warning letters and noting that no probiotic is approved as a drug for infants of any age. Broader questions about supplement quality and manufacturing are covered in the piece on supplement safety.
Where fermented foods and fiber fit
The most interesting recent finding in this area did not involve a supplement. Wastyk and colleagues, publishing in Cell in 2021, randomized 36 healthy adults to a high-fermented-food diet or a high-fiber diet for ten weeks. In the fermented food arm — yogurt, kefir, kimchi, kombucha, fermented vegetables — overall microbial diversity increased, with larger servings producing larger effects, and 19 inflammatory proteins in blood decreased, including interleukin-6. The high-fiber arm did not show those changes over the same period.
That is one small trial and should be read as such. But it points at something the supplement framing misses: a fermented food delivers a large and varied microbial load in a food matrix, repeatedly, as part of eating. Fiber matters over longer timescales by feeding the organisms already resident — the subject of the article on apple pectin and soluble fiber.
What the previous version of this post claimed
Several claims stood here that we retract. The post said specific strains “may produce a natural antibiotic effect” and help prevent infections by E. coli, salmonella, clostridia, staphylococci and candida. That is a disease claim, it was not supported by the sources cited, and it should never have been published.
It said probiotic cultures contribute to the manufacture of B vitamins “such as B12.” Colonic bacteria do synthesize cobalamin, but humans cannot absorb it there. As set out in PLOS Pathogens, absorption occurs in the small intestine, upstream of where colonic synthesis happens, so that B12 is unavailable to the host. Anyone relying on gut bacteria for B12 is not getting any.
It also cited a 1984 study of fructo-oligosaccharides and blood glucose in people with diabetes, which is a disease claim and not something a food ingredient should be marketed on. Its supporting references were trade publications and newsletters from the 1980s and 1990s rather than peer-reviewed clinical evidence. Whether nutrient source and form matter is handled in natural versus synthetic nutrients.
Greens Plus does include probiotic cultures, and our labels list what is in each serving. Any product’s cultures should be judged against the strain designations and doses used in published trials, which is the standard we would ask readers to apply to ours as much as anyone’s.
Common questions
Should I take a probiotic while on antibiotics?
The meta-analytic evidence for reducing antibiotic-associated diarrhea is the strongest in this field, and the NIH notes that starting a studied strain within two days of the first antibiotic dose is what the trials did. Benefit appears concentrated in people at higher baseline risk. It is worth asking the prescriber, particularly if you are immunocompromised or seriously ill.
Do I need a probiotic if my digestion is fine?
No evidence supports routine use in healthy adults. The Genome Medicine review found no measurable change in fecal microbiota composition in this group, and no guideline body recommends supplementation without a specific indication.
Is a higher CFU count better?
Not in itself. The NIH states directly that higher-CFU products are not necessarily more effective. What matters is whether the specific strain was studied at that dose, and whether the count is guaranteed at the end of shelf life rather than at the moment of manufacture.
Are probiotics safe for everyone?
Not everyone. The evidence above identifies real risk in critically ill patients, immunocompromised people, those with central venous catheters, and hospitalized preterm infants. Anyone in those categories should treat probiotics as a medical decision.
The bottom line
Probiotics have a small number of well-supported uses tied to named strains at studied doses, and a marketing footprint that extends far past them. Reading the strain designation, checking the dose against the trial and ignoring the CFU headline gets you most of the way to a sensible decision. How diet relates to normal immune function is worth reading alongside this one.
Greens Plus formulates whole-food blends with full disclosure of what each serving contains, so the label math is available to anyone who wants to do it. The range is at greensplus.com.
About the Author
